Our long-term goal is to create computational brain simulations sophisticated enough for use in drug testing. This goal is relevant to all houses of NIH seeking pharmacological cures for cognitive, emotional, or behavioral pathologies including addictions and depressions. Drug testing by computer simulations will decrease the cost of drug development and shorten the time required to develop such therapies. It will also help eliminate treatments with undesirable effects on cognition, emotion, and behavior. Using appropriate data sets for studying hippocampal function, we have proof of concept that motivates moving toward this long-term goal. Because neurobiologically-based simulations can predict certain aspects of behavior and cognition, it is time to extend the purview of such simulations to more brain systems and to make such simulations relevant to neuroactive drugs. This application focuses on a particular pathology, severe, acute post-traumatic stress disorder (PTSD). PTSD is distinct from depressions treated by the classic antidepressants and selective serotonin reuptake inhibitors (SSRI's). This application presents a novel, integrated brain theory of PTSD and emphasizes the critical nature of sleep. PTSD not only correlates with poor sleep, but nightmares are the most common symptom in severe, acute PTSD. Experimental animal models already produce data on the effects of trauma on sleep and dreaming; therefore, challenges for the proposed technology are (1) to build a computational model that reproduces such experimental results and (2) to predict novel therapeutic formulations for speeding the improvement of sleep quality, especially dream sleep (REM). The computer simulations will account for interactions at the level of neurons, at the level of synapses, and at the level of drug-receptor interactions. As part of our long-term goal, such software will be able to predict: (1) the effect of gene doses and (2) the effect of behavioral experiences on sleep. Extension of the existing computational models includes additional brain regions and more receptor systems. Moreover, the new software must interface two different styles of brain modeling: one corresponding to the biochemistry and pharmacology of sleep, dreaming, and certain aspects of stress that collapses neurons and synapses into scalar interactions versus the other which corresponds to models and simulations that reproduce contextual and episodic learning by forebrain cortical systems using tens of thousands of neurons and billions of synapses. Specifically, the cognition-mediating system consisting of the hippocampal formation, the basolateral amygdala and neocortex must be interfaced with each other and with the sleep and stress systems, including the n. centralis amygdala (CeA) and several small nuclei located in the brainstem and hypothalamus. Importantly, even if our theory of PTSD is wrong, the software being proposed here has use for modeling competing theories of sleep disruption. [unreadable] [unreadable] Our long-term goal is to create computational brain simulations that are sophisticated enough for drug testing; this goal is relevant to all houses of NIH seeking pharmacological cures for cognitive, emotional, or behavioral pathologies including addictions and depressions. Drug testing by computer simulations will decrease the cost of drug development and shorten the time required to develop such therapies, and it will also help eliminate treatments with undesirable effects on cognition, emotion, and behavior. [unreadable] [unreadable] [unreadable]